Safran Builds Emergency Locator Beacon SDR Receiver Using Model-Based Design and Analog Devices System on Module Hardware
Develop receiver hardware for emergency locator beacons within a tight deadline
Adopt Model-Based Design with MATLAB and Simulink to design a specialized SDR and implement it on an Analog Devices RF system on module
- Analysis and testing times shortened by eight months
- FPGA implementation time reduced by at least 50%
- Adoption of Model-Based Design jumpstarted
Emergency locator beacons make it possible to quickly locate aircraft, ships, and even individuals in distress. Once activated, the beacon transmits a 406 MHz radio signal to satellites in the international Cospas-Sarsat system, which relays the beacon’s position to ground stations. The beacon can also transmit signals directly to search-and-rescue teams equipped with beacon monitoring devices.
To increase localization accuracy and reduce power consumption, second-generation beacons (SGBs) will use spread-spectrum technology and offset quadrature phase-shift keying (OQPSK). Safran, formerly Orolia, has built a prototype software-defined radio (SDR) that receives, detects, processes, and decodes SGB signals. Safran engineers used Model-Based Design with MATLAB® and Simulink® to model and simulate the receiver and worked with MathWorks engineers to implement it on an Analog Devices® SDR system on module (SoM).
“Our Simulink simulations enabled us to validate most elements of the design, such as demodulation and decoding, well before involving hardware,” says Boris Van Amerongen, senior director of R&D at Safran. “This approach gave us a high level of confidence in our reception algorithms and enabled us to deliver a full prototype in just 12 months.”
SGB signals use significantly more complex spread-spectrum modulation schemes than first-generation beacons. The Safran team recognized that their legacy analog tools would be inadequate for testing SGB signal transmissions. In addition, Safran engineers had worked primarily with analog transmitter designs, and had little experience with designing and implementing a digital receiver.
To meet production goals, the company set an aggressive project deadline: the team was to deliver a functional prototype in just one year. The team wanted to implement the receiver as an SDR that could be adapted or modified to meet customers’ requirements. Lacking engineers with experience in FPGA integration, the team needed a way to implement the SDR design on the target SoM.
Safran used Model-Based Design with MATLAB and Simulink to accelerate the development of an SDR receiver for second-generation emergency locator beacons.
In MATLAB, the team analyzed and visualized in-phase quadrature (IQ) data from SGB signals acquired via a Tektronix® spectrum analyzer with Instrument Control Toolbox™. They used the analysis results to characterize chip rate and other signal parameters and to guide the receiver design.
Working in Simulink with Communications Toolbox™, the team modeled the receive chain, including modules for Bose–Chaudhuri–Hocquenghem (BCH) decoding, demodulation, frequency offset estimation, despreading, and synchronization. They ran simulations to test each individual module and the complete chain. Using the Communications Toolbox Support Package for Xilinx® Zynq®-Based Radio, the team then validated the receiver design with real-world RF signals captured via a Zynq SoC.
The Safran team worked with MathWorks engineers to implement the design on the Analog Devices SDR SoM target. The two teams converted the initial floating-point model to fixed point using Fixed-Point Designer™. They generated synthesizable HDL for the SoC programmable logic with HDL Coder™. They used Embedded Coder® to generate C code for the BCH decoder and other elements of the design that were better suited to implementation on the SoC’s ARM processor.
The team verified synchronization and other aspects of the design using HDL Verifier™ and conducted reduced-range indoor tests as well as field tests to demonstrate the SDR’s ability to process SGB signals from emergency locator beacon transmitters.
- Analysis and testing times shortened by eight months. “Starting from scratch, we were able to analyze and test complex SGB signals in about four months,” says Van Amerongen. “Without MATLAB, it would have taken us three times longer. This turnaround time was key because it enabled us to kickstart development of our receiver chain.”
- FPGA implementation time reduced by at least 50%. “Writing the HDL code ourselves would have doubled, if not tripled, the amount of time we needed to implement our design on the FPGA,” Van Amerongen notes. “Generating the HDL from our Simulink model with HDL Coder was not only faster, it also enabled us to complete the project without involving an FPGA specialist—a resource that we didn’t have at the time.”
- Adoption of Model-Based Design jumpstarted. “The support we received from MathWorks was exceptional,” says Van Amerongen. “Model-Based Design was new to us, as was SDR and receiver design. The availability and responsiveness of MathWorks support enabled us to progress quickly right from the start.”